This study examines male and female hazard rates in the periods 1968-1971 and 1979-1982 using data for young workers from the various samples of the National Longitudinal Surveys. Contrary to a number of previous micro-data studies, I demonstrate that for the period 1968-1971 female workers quit their initial full-time jobs at substantially higher rates than male workers. Moreover, while male hazard rates show a monotonic decline, female rates show a nonmonotonic u-shaped pattern, which I attribute to a "birth effect" -- young women leaving the labor force to have children.

For the period 1979-1982, however, young women had become almost indistinguishable from young men in terms of job tenure, attachment to the labor force, and percentage of workers who are professional, managerial, and technical. The finding of the equality in hazard rates between male and female workers in the later period was invariant to different parametric assumptions about the nature of duration dependence and the existence of unobserved heterogeneity.

Two factors contributed to the elimination of the first-job "tenure gap" between young men and women: (1) women' s increased commitment to the paid workforce, and (2) their increasing age at the time of first marriage and/or first pregnancy. Evidence from examining the last job held during the sample period suggests that these factors delay, but do not entirely eradicate, the point at which women begin to leave their jobs at a higher rate than men.

In the period 1968-1971 the female-male ratio of expected tenure on initial full-time jobs was 59% and the corresponding ratio of earnings was roughly 73%. By 1979-1982, the tenure gap had closed and the earnings gap had narrowed to almost 90%. Since the narrowing of the wage gap seems to lag the narrowing of the tenure gap, the direction of causation !)lay be from lower tenure to lower wages.